A real-time UAV scheduling strategy control method

Abstract

The present application provides a real-time drone scheduling strategy control method based on task information and constraints, including: obtaining the initial path planning result of the multi-objective delivery points of the drone, and establishing an initial mapping relationship between the pheromone matrix and the position matrix; injecting a weight coefficient into the pheromone matrix according to the priority information of the multi-objective delivery points, and increasing the enhancement amplitude of the pheromone concentration of the surrounding paths if a high-priority delivery point is detected; if the wind speed or rainfall intensity in the monitored weather change data exceeds the preset range, updating the concentration distribution of the pheromone matrix with a dynamically adjusted evaporation rate; if the concentration value of the pheromone matrix is lower than the preset threshold and the position update direction deviates from the target path, re-initializing the pheromone concentration distribution in this area; generating a final path planning scheme according to the updated pheromone matrix and the position matrix.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular, to a real-time unmanned aerial vehicle (UAV) scheduling strategy control method based on task information and constraints. Background Art

[0002] In the scheduling problem of multiple UAV target delivery points in a complex environment, there are multiple delivery points and dynamic environmental constraints, and it is necessary to optimize both the delivery path and time efficiency simultaneously. When dealing with such problems, traditional methods often have difficulty in balancing global search ability and local fine adjustment. Combining the pheromone update mechanism of the ant colony algorithm with the position update of the particle swarm optimization algorithm attempts to solve this contradiction, but technical problems still exist in practical applications.

[0003] The pheromone update mechanism of the ant colony algorithm depends on the accumulation of pheromone on the path, and guides the search direction through the evaporation and enhancement of pheromone. However, in a dynamic environment, the UAV needs to adjust the path in real time to avoid obstacles or respond to emergencies. It is difficult for the update frequency and intensity of pheromone to adapt to such rapid changes, which may lead to the search falling into a local optimum or being unable to respond to environmental changes in a timely manner. Summary of the Invention

[0004] The present invention provides a real-time UAV scheduling strategy control method based on task information and constraints, mainly including:

[0005] Obtain the initial path planning result of the UAV multiple target delivery points, and establish an initial mapping relationship between the pheromone matrix and the position matrix; if the position information of dynamic obstacles is detected in the initial path planning result, reduce the concentration value of the pheromone matrix corresponding to this path, and increase the concentration value of the pheromone matrix in the adjacent area; according to the current concentration value of the pheromone matrix, use the particle swarm optimization algorithm to adjust the position update direction of the UAV, and if the concentration value is higher than the preset threshold, preferentially select this direction; inject a weight coefficient into the pheromone matrix according to the priority information of the multiple target delivery points, and if a high-priority delivery point is detected, increase the enhancement amplitude of the pheromone concentration of its surrounding paths; if the wind speed or rainfall intensity in the weather change data is detected to exceed the preset range, update the concentration distribution of the pheromone matrix using the dynamically adjusted evaporation rate; if the concentration value of the pheromone matrix is lower than the preset threshold and the position update direction deviates from the target path, re-initialize the concentration distribution of the pheromone in this area; generate the final path planning scheme according to the updated pheromone matrix and the position matrix.

[0006] The technical solution provided by the embodiments of the present invention may include the following beneficial effects: The real-time UAV scheduling strategy control method based on task information and constraints provided by the present invention can construct an initial mapping relationship between the pheromone matrix and the position matrix by obtaining the initial path planning results of the multi-target delivery points of the UAV, providing a basic data structure for path optimization. Further, the present invention reduces the pheromone concentration of the corresponding path by detecting dynamic obstacles in the initial path and increases the pheromone concentration in the adjacent area to guide the UAV to avoid obstacles. Further, the present invention also adjusts the UAV position update direction according to the pheromone concentration by using the particle swarm optimization algorithm, preferentially selects the high-concentration value direction, and balances the global search and local optimization. Finally, the present invention also injects a weight coefficient according to the priority of the delivery point, and the enhancement amplitude of the pheromone concentration of the path around the high-priority delivery point increases to ensure that important tasks are completed first. In addition, when the present invention monitors that the wind speed or rainfall intensity exceeds the preset range, it dynamically adjusts the pheromone evaporation rate, updates the concentration distribution, and adapts to the impact of weather changes on flight. When the pheromone concentration is lower than the threshold and the position update direction deviates from the target path, it re-initializes the pheromone concentration distribution of the area to avoid local optimum and path deviation. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 It is a flowchart of a real-time UAV scheduling strategy control method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0008] To further understand the content of the present invention, the present invention will be described in detail with reference to the drawings and embodiments. The following will further elaborate on the present application with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related invention and are not intended to limit the invention. In addition, it should be noted that for the sake of description, only the parts related to the invention are shown in the drawings.

[0009] As Figure 1 , a real-time UAV scheduling strategy control method based on task information and constraints in this embodiment may specifically include:

[0010] S101. Obtain the initial path planning results of the multi-target delivery points of the UAV, and establish an initial mapping relationship between the pheromone matrix and the position matrix.

[0011] Obtain the initial path planning result of the multi-target delivery points of the UAV, and extract the coordinate information of the path nodes. For the coordinate information of the path nodes, construct a position matrix to store the longitude and latitude data of each node. According to the number of nodes in the position matrix, initialize the pheromone matrix and set the matrix structure with the same dimension. Traverse each node in the position matrix, calculate the Euclidean distance between the nodes, and obtain the distance matrix. Normalize each element value in the distance matrix and convert it into the initial pheromone concentration value. Fill the corresponding positions of the pheromone matrix according to the initial pheromone concentration value to complete the matrix initialization. Establish the mapping relationship between the pheromone matrix and the position matrix, and realize the data association between the matrices through node indexing.

[0012] Specifically, the path planning of the multi-target delivery points of the UAV is a complex task that needs to consider multiple factors and be optimized. The initial path planning result is usually obtained based on simple heuristic algorithms, such as the nearest neighbor method or the greedy algorithm. These algorithms can quickly generate an initial solution, but they are usually not globally optimal. Extracting the coordinate information of the path nodes is to prepare for subsequent optimization. Each node represents a delivery point or a transfer station, and its coordinates are usually represented by longitude and latitude. Constructing a position matrix can conveniently store and access this coordinate data, providing a basis for subsequent distance calculation and path optimization. The pheromone matrix is a key concept in the ant colony algorithm, which is used to simulate the chemical information traces left by ants in nature. When initializing the pheromone matrix, usually set all elements to the same small positive number to ensure that the probability of each path being selected is equal at the beginning of the algorithm. This can avoid the algorithm converging to a local optimal solution prematurely. Calculating the Euclidean distance between the nodes is to construct the distance matrix. In one embodiment, the following formula can be used to calculate the spherical distance between two points. Specifically:

[0013] , where d represents the spherical distance between the nodes, R represents the radius of the earth, φ1 and φ2 represent the latitudes of the two points, and λ1 and λ2 represent the longitudes of the two points.

[0014] The distance matrix provides an important reference for subsequent path selection. Normalizing the distance values and converting them into the initial pheromone concentration values is a commonly used technique. By doing so, the initial pheromone distribution is inversely proportional to the actual distance, that is, the closer the distance, the higher the initial pheromone concentration. This setting can help the algorithm converge to a better solution more quickly. Establishing the mapping relationship between the pheromone matrix and the position matrix is to facilitate coordinate lookup and path construction during the optimization process. Through the node index, it is possible to quickly switch between the two matrices, improving the efficiency of the algorithm. For example, suppose there is a scenario with 5 delivery points. The initial path may be directly connected in the input order: A→B→C→D→E. Based on the coordinates of A→B→C→D→E, the distance matrix can be calculated. After normalizing these distances, the initial pheromone concentration can be obtained. For instance, if the shortest distance is set to 1 and other distances are scaled proportionally, then the pheromone concentration may show a distribution with high values for short distances and low values for long distances. This initialization method provides a good starting point for subsequent path optimization. The ant colony algorithm can iteratively search based on this, gradually adjusting the pheromone distribution, and finally finding a better delivery path. The optimized path may be in a sequence such as A→E→C→B→D, which may significantly reduce the total distance compared to the initial path. Through this method, not only geographical location information is considered, but also the necessary data structure is prepared for the intelligent optimization algorithm. The advantage of this method is that it combines deterministic distance information and a random search strategy, and can find a solution close to the optimal one within a reasonable time.

[0015] S102. If the dynamic obstacle position information exists in the initial path planning result, reduce the pheromone matrix concentration value corresponding to this path, and increase the pheromone matrix concentration value in the adjacent area.

[0016] Obtain the initial path planning result and extract the dynamic obstacle position information therein. According to the dynamic obstacle position information, judge the index value of the path where it is located. For the path corresponding to the index value, reduce the pheromone concentration value of this path in the pheromone matrix. According to the dynamic obstacle position information, determine the range of its adjacent area. For the paths in the adjacent area, increase the pheromone concentration value of the corresponding paths in the pheromone matrix. Update the pheromone matrix and save the adjusted pheromone concentration value. Based on the updated pheromone matrix, recalculate the priority of path planning.

[0017] Specifically, during the path planning process, when dynamic obstacle position information is detected in the initial path planning result, the system will first calculate the influence range of the obstacle on the path based on the position and movement trajectory of the obstacle. Assuming the obstacle is located at coordinates (x = 10, y = 20) with an influence radius of 5 meters, the system will reduce the pheromone matrix concentration value corresponding to this path from the initial value of 10 to 6 to reduce the attractiveness of this path. At the same time, the system will increase the pheromone concentration value in the adjacent area. For example, at coordinates (x = 12, y = 22), the pheromone concentration value will be increased from 0 to 4 to guide the path planning algorithm to select a safer path. In other embodiments, this process is implemented through the ant colony algorithm, and the pheromone update formula in the algorithm is:

[0018]

[0019] τ ij represents the path pheromone concentration, ρ represents the pheromone evaporation coefficient, Δτ ij represents the pheromone increment, which is dynamically adjusted according to the obstacle influence range and path safety level, and t represents the current time. The system will analyze the movement trajectory of the obstacle in real time, predict its future position, and update the pheromone matrix to ensure the real-time and safety of path planning. In this way, the system can achieve efficient path planning in a dynamic environment and avoid collisions with obstacles.

[0020] S103. According to the current concentration value of the pheromone matrix, use the particle swarm optimization algorithm to adjust the position update direction of the UAV. If the concentration value is higher than the preset threshold, preferentially select this direction.

[0021] Obtain the concentration values of all paths in the pheromone matrix and determine whether they are higher than the preset threshold. If the concentration value is higher than the preset threshold, use the particle swarm optimization algorithm to calculate the UAV position update direction. According to the result of the particle swarm optimization algorithm, determine the priority of the UAV position update direction. For the direction with the highest priority, adjust the UAV position update path. Update the concentration value of the corresponding path in the pheromone matrix. According to the updated pheromone matrix, recalculate the UAV path planning result. Save the adjusted UAV path planning result.

[0022] Specifically, in the current pheromone matrix, assume that the pheromone concentrations at the position where the UAV is located are [8, 6, 9, 7] respectively, and the preset threshold is 7. When using the particle swarm optimization algorithm to adjust the UAV's position update direction, first calculate the difference between the pheromone concentration of each direction and the threshold, obtaining [1, -1, 2, 0]. According to the update formula of the particle swarm optimization algorithm, the UAV's speed is updated as: v = w * v + c1 * rand() * (p best - x) + c2 * rand() * (g best- x), where w is the inertia weight set to 5, c1 and c2 are learning factors set to 5 and 5 respectively, rand() is the random number generation function, p best , g best , representing the individual optimal position and the global optimal position respectively.

[0023] In one embodiment, assume the current position of the drone is [0, 0], the speed is [5, 5], the individual optimal position is [8, 9], and the global optimal position is [7, 8]. Through calculation, the new speed v x = 5 * 5 + 5 * 3 * (8 - 0) + 5 * 4 * (7 - 0) = 25 - 09 - 18 = -02, v y = 5 * 5 + 5 * 2 * (9 - 0) + 5 * 3 * (8 - 0) = 25 - 03 - 09 = 13. The updated position is x = x + v, that is, [0 + (-02), 0 + 13] = [98, 13]. Since the direction with the highest pheromone concentration is 9, this direction is preferentially selected for position update, and finally the position of the drone is adjusted to [98, 13] to ensure that it explores and executes tasks in the area with high pheromone concentration.

[0024] S104. Inject a weight coefficient into the pheromone matrix according to the priority information of the multi-objective delivery points. If a high-priority delivery point is detected, increase the enhancement amplitude of the pheromone concentration of its surrounding paths.

[0025] Obtain the priority information of the delivery points, and assign weight coefficients to each delivery point according to the priority level. Inject the weight coefficients into the pheromone matrix to adjust the pheromone concentration values of the corresponding paths in the matrix. Use a detection algorithm to identify high-priority delivery points and determine the surrounding path range. For the surrounding paths of high-priority delivery points, calculate the enhancement amplitude of the pheromone concentration. Adjust the concentration values of the corresponding paths in the pheromone matrix according to the enhancement amplitude, and update the matrix data. Use a path planning algorithm to calculate the drone flight path according to the updated pheromone matrix. Determine the final drone flight path and save the path planning result.

[0026] The weight coefficients assigned to each delivery point can be obtained through the following formula:

[0027]

[0028] Where, W i represents the weight coefficient of the delivery point, P i represents the priority value of the delivery point, α i represents the influence factor, β i represents the adjustment coefficient, and n represents the total number of delivery points.

[0029] Injecting the weight coefficient into the pheromone matrix and adjusting the pheromone concentration value of the corresponding path in the matrix is specifically obtained through the following formula:

[0030]

[0031] τ ij represents the pheromone concentration of the path, ρ represents the pheromone evaporation coefficient, and Δτ ij represents the pheromone increment, W p represents the weight adjustment value, and t represents the current time.

[0032] Furthermore, the influence range of high-priority delivery points can be calculated through the following formula:

[0033]

[0034] Among them, R p represents the influence range of high-priority delivery points, γ d represents the distance attenuation coefficient, d represents the distance value, σ represents the standard deviation, and D represents the maximum influence distance.

[0035] Moreover, the pheromone enhancement amount is affected by the influence range of the high-priority delivery points and can be calculated through the following formula:

[0036]

[0037] Among them, Δτ represents the pheromone enhancement amount, λ represents the enhancement coefficient, Q represents the pheromone constant, L represents the path length, and ω represents the weight coefficient.

[0038] Finally, calculate the path selection probability according to the formula:

[0039] ;

[0040] P ij represents the path selection probability, τ ij represents the pheromone concentration, η ij represents the heuristic information, α represents the importance coefficient of the pheromone, β represents the importance coefficient of the heuristic factor, and N i represents the set of selectable paths. High-probability paths or paths greater than a certain threshold can be selected to determine the final flight path of the drone.

[0041] S105. If the wind speed or rainfall intensity in the monitored weather change data exceeds the preset range, update the concentration distribution of the pheromone matrix with the dynamically adjusted evaporation rate.

[0042] Obtain the wind speed value and rainfall value of real-time monitoring, and compare them with the preset wind speed threshold and rainfall threshold. If the wind speed value or rainfall value exceeds the preset threshold, start the dynamic adjustment mechanism to calculate the evaporation rate value. According to the dynamically adjusted evaporation rate value, determine the update range of the concentration distribution in the pheromone matrix. Reassign the concentration values in the pheromone matrix using the update range to obtain the adjusted pheromone matrix. Use the adjusted pheromone matrix as the input, and combine with the ant colony algorithm to perform path optimization calculation. Based on the optimized calculation results, generate a new path planning scheme. Compare the new path planning scheme with the original scheme and output the final path selection result.

[0043] Specifically, after the wind speed in the monitored weather change data exceeds the preset 10 m / s or the rainfall intensity exceeds the preset 50 mm / h, the system will start the dynamic adjustment mechanism. First, collect real-time data through the wind speed sensor and rain gauge. The wind speed sensor collects data once per second, and the rain gauge collects data once per minute, and input these data into the data processing module. The data processing module uses the Kalman filtering algorithm to smooth the collected data to reduce noise interference. For example, when the wind speed is 12 m / s, the Kalman filtering algorithm will smooth this outlier to 15 m / s to ensure the stability of the data. Then, the system will calculate the evaporation rate using the dynamic adjustment algorithm based on the smoothed data. Specifically, when the wind speed is 15 m / s, the evaporation rate is adjusted to 2 times the original; when the rainfall intensity is 60 mm / h, the evaporation rate is adjusted to 5 times the original. The adjusted evaporation rate will immediately update the concentration distribution of the pheromone matrix. The pheromone matrix uses a Gaussian distribution model, and by calculating the concentration change of each node, dynamically adjusts the diffusion range of the pheromone. For example, when the evaporation rate is adjusted to 2 times, the concentration of each node in the pheromone matrix will be updated according to the formula of the Gaussian distribution, and the new concentration distribution will reflect the influence of the wind speed and rainfall intensity on the pheromone diffusion. Finally, the system will store the updated pheromone matrix in the database and push it to the relevant application modules in real time for subsequent path planning or decision support systems to use. The whole process is realized through an automated process without manual intervention, ensuring the real-time and accuracy of data processing.

[0044] S106. If the concentration value of the pheromone matrix is lower than the preset threshold and the position update direction deviates from the target path, re-initialize the pheromone concentration distribution in this area.

[0045] Obtain the concentration value of the pheromone matrix according to the preset threshold, and determine whether the concentration value is lower than the threshold. If the concentration value is lower than the threshold, obtain the direction value of position update, and determine whether the direction value deviates from the target path. If the direction value deviates from the target path, determine the pheromone distribution in this area, and use the initialization algorithm to regenerate the pheromone distribution. Obtain the re-initialized pheromone matrix, and use the ant colony algorithm to update the path planning. According to the updated path planning, use the genetic algorithm to optimize the target path. Through the optimized target path, obtain the final pheromone distribution matrix.

[0046] For example, in the ant colony optimization algorithm, if the concentration value of the pheromone matrix is lower than the preset threshold 1 and the deviation angle between the position update direction and the target path exceeds 15 degrees, the system will automatically trigger the re-initialization of the pheromone concentration in this area. Specifically, the system first evaluates the deviation degree by calculating the cosine similarity between the current position and the target path. If the similarity is lower than 9, it is determined to be a deviation. Subsequently, the system uses the Gaussian distribution function to regenerate the pheromone concentration in this area, with the mean set to 5 and the standard deviation to 1, to ensure a more uniform distribution of pheromones. At the same time, the system will use the optimal path information in the historical data to adjust the newly generated pheromone concentration by the weighted average method, making the new distribution more inclined to guide the ants towards the target path. This process is achieved through matrix operations to ensure computational efficiency and accuracy. Through this dynamic adjustment mechanism, the system can effectively handle path planning problems in complex environments and improve the stability and convergence speed of the algorithm.

[0047] S107. Generate a final path planning scheme according to the updated pheromone matrix and the position matrix.

[0048] Obtain the updated pheromone matrix and position matrix. If the pheromone value of a certain node in the pheromone matrix is higher than the preset threshold, determine that node as a candidate node. Extract the candidate node coordinates from the position matrix, group the candidate nodes using the clustering algorithm to obtain the node clustering result. For the node clustering result, calculate the center point coordinates of each group of nodes to generate a set of center points. According to the set of center points, use the shortest path algorithm to calculate the optimal path between the center points to obtain the preliminary path planning result. If there are overlapping paths in the preliminary path planning result, adjust the node order of the overlapping paths to optimize the path planning result. Through the optimized path planning result, generate a final path planning scheme to determine the node access order and path direction. Match the final path planning scheme with the position matrix and output the detailed node coordinates and path information of the path planning scheme.

[0049] The above content is only an example and explanation of the structure of the present invention. Those skilled in the art of this technology can make various modifications or supplements to the described specific embodiments or use similar methods for substitution. As long as they do not deviate from the structure of the invention or exceed the scope defined by this claim book, they should all fall within the protection scope of the present invention.

Claims

1. A real-time UAV scheduling strategy control method based on task information and constraints, characterized in that, The method includes: Obtain the initial path planning result of the multi-target delivery points of the unmanned aerial vehicle (UAV), and establish the initial mapping relationship between the pheromone matrix and the position matrix; if the position information of dynamic obstacles is detected in the initial path planning result, reduce the pheromone matrix concentration value corresponding to this path, and increase the pheromone matrix concentration value in the adjacent area; according to the current concentration value of the pheromone matrix, use the particle swarm optimization algorithm to adjust the position update direction of the UAV, and if the concentration value is higher than the preset threshold, preferentially select this direction; inject the weight coefficient into the pheromone matrix according to the priority information of the multi-target delivery points, and if a high-priority delivery point is detected, increase the pheromone concentration enhancement amplitude of the surrounding paths, specifically including: obtain the priority information of the delivery points, and assign weight coefficients to each delivery point according to the priority level. The weight coefficient assigned to each delivery point is obtained through the following formula: Among them, W i represents the weight coefficient of the distribution point, P i represents the priority value of the distribution point, α i represents the influence factor, β i represents the adjustment coefficient, and n represents the total number of distribution points Inject the weight coefficient into the pheromone matrix, and adjust the pheromone concentration value of the corresponding path in the matrix, which is specifically obtained through the following formula: τ ij represents the concentration of path pheromone, ρ represents the pheromone evaporation coefficient, and Δτ ij represents the pheromone increment, and W p represents the weight coefficient, and t represents the current time Use the detection algorithm to identify high-priority delivery points and determine the surrounding path range, specifically including: calculate the influence range of high-priority delivery points through the following formula: Among them, R p represents the influence range of high-priority delivery points, and γ d represents the distance attenuation coefficient, d represents the distance value, σ represents the standard deviation, and D represents the maximum influence distance For the surrounding paths of high-priority delivery points, calculate the enhancement amplitude of the pheromone concentration. The pheromone enhancement amount is affected by the influence range of the high-priority delivery point, and is calculated through the following formula: where Δτ represents the pheromone enhancement amount, λ represents the enhancement coefficient, Q represents the pheromone constant, L represents the path length, and ω represents the weight coefficient. Adjust the concentration value of the corresponding path in the pheromone matrix according to the enhancement amplitude, and update the matrix data. Use the path planning algorithm to calculate the UAV flight path according to the updated pheromone matrix, specifically including: calculate the path selection probability according to the formula: ， P ij represents the path selection probability, τ ij represents the pheromone concentration, η ij represents the heuristic information, α represents the importance coefficient of pheromone, β represents the importance coefficient of heuristic factor, N i represents the set of selectable paths Determine the final UAV flight path and save the path planning result; if the wind speed or rainfall intensity in the weather change data is detected to exceed the preset range, update the concentration distribution of the pheromone matrix with the dynamically adjusted evaporation rate; if the concentration value of the pheromone matrix is lower than the preset threshold and the position update direction deviates from the target path, re-initialize the pheromone concentration distribution in this area; generate the final path planning scheme according to the updated pheromone matrix and the position matrix.

2. The method according to claim 1, wherein The obtaining of the initial path planning result of the multi-target delivery points of the UAV and the establishment of the initial mapping relationship between the pheromone matrix and the position matrix include: Obtain the initial path planning result of the multi-target delivery points of the UAV, and extract the path node coordinate information. For the path node coordinate information, construct a position matrix to store the latitude and longitude data of each node. According to the number of nodes in the position matrix, initialize the pheromone matrix and set the matrix structure with the same dimension. Traverse each node in the position matrix, calculate the Euclidean distance between nodes, and obtain the distance matrix. Normalize each element value in the distance matrix and convert it into the initial pheromone concentration value. Fill the corresponding positions of the pheromone matrix according to the initial pheromone concentration value to complete the matrix initialization. Establish the mapping relationship between the pheromone matrix and the position matrix, and realize the data association between the matrices through node indexing.

3. The method according to claim 1, wherein If it is detected that there is dynamic obstacle position information in the initial path planning result, the pheromone matrix concentration value corresponding to this path is reduced, and the pheromone matrix concentration value is increased in the adjacent area, including: Obtain the initial path planning result and extract the dynamic obstacle position information therein; According to the dynamic obstacle position information, judge the index value of the path where it is located; For the path corresponding to the index value, reduce the pheromone concentration value of this path in the pheromone matrix; According to the dynamic obstacle position information, determine the range of its adjacent area; For the paths in the adjacent area, increase the pheromone concentration value of the corresponding paths in the pheromone matrix; Update the pheromone matrix and save the adjusted pheromone concentration value; Based on the updated pheromone matrix, recalculate the priority of path planning.

4. The method according to claim 1, wherein According to the current concentration value of the pheromone matrix, use the particle swarm optimization algorithm to adjust the position update direction of the UAV. If the concentration value is higher than the preset threshold, this direction is preferentially selected, including: Obtain the concentration values of all paths in the pheromone matrix and judge whether they are higher than the preset threshold; If the concentration value is higher than the preset threshold, use the particle swarm optimization algorithm to calculate the UAV position update direction; According to the result of the particle swarm optimization algorithm, determine the priority of the UAV position update direction; For the direction with the highest priority, adjust the UAV position update path; Update the concentration value of the corresponding path in the pheromone matrix; According to the updated pheromone matrix, recalculate the UAV path planning result; Save the adjusted UAV path planning result.

5. The method according to claim 1, wherein If it is monitored that the wind speed or rainfall intensity in the weather change data exceeds the preset range, the concentration distribution of the pheromone matrix is updated using the dynamically adjusted evaporation rate, including: Obtain the real-time monitored wind speed value and rainfall value, and compare them with the preset wind speed threshold and rainfall threshold; If the wind speed value or rainfall value exceeds the preset threshold, start the dynamic adjustment mechanism to calculate the evaporation rate value; According to the dynamically adjusted evaporation rate value, determine the update amplitude of the concentration distribution in the pheromone matrix; Re-distribute the concentration values in the pheromone matrix using the update amplitude to obtain the adjusted pheromone matrix; Use the adjusted pheromone matrix as the input and combine the ant colony algorithm to perform path optimization calculation; Based on the optimization calculation result, generate a new path planning scheme; Compare the new path planning scheme with the original scheme and output the final path selection result.

6. The method according to claim 1, characterized in that If the concentration value of the pheromone matrix is lower than the preset threshold and the position update direction deviates from the target path, re-initialize the pheromone concentration distribution in this area, including: Obtain the concentration value of the pheromone matrix according to the preset threshold and judge whether the concentration value is lower than the threshold; If the concentration value is lower than the threshold, obtain the direction value of the position update and judge whether the direction value deviates from the target path; If the direction value deviates from the target path, determine the pheromone distribution in this area and use the initialization algorithm to regenerate the pheromone distribution; Obtain the re-initialized pheromone matrix and use the ant colony algorithm to update the path planning; According to the updated path planning, use the genetic algorithm to optimize the target path; Through the optimized target path, obtain the final pheromone distribution matrix.

7. The method according to claim 1, characterized in that Generating a final path planning scheme based on the updated pheromone matrix and the position matrix includes: Obtain the updated pheromone matrix and position matrix. If the pheromone value of a certain node in the pheromone matrix is higher than the preset threshold, determine that node as a candidate node; Extract the candidate node coordinates from the position matrix, and use a clustering algorithm to group the candidate nodes to obtain the node clustering result; For the node clustering result, calculate the center point coordinates of each group of nodes to generate a set of center points; According to the set of center points, use the shortest path algorithm to calculate the optimal path between the center points to obtain the preliminary path planning result; If there is path overlap in the preliminary path planning result, adjust the node order of the overlapping paths to optimize the path planning result; Based on the optimized path planning result, generate a final path planning scheme to determine the node access order and path direction; Match the final path planning scheme with the position matrix and output the detailed node coordinates and path information of the path planning scheme.

Images